Thoughts on Studio Automation with Marc Berenson

Key Takeaways

• Elevate the right people above the process, to assess the workflow, tools involved, and any existing scripts or automations being used ad hoc.
• Build your automation like bricks in a wall. Make a good foundation and build from there.
• It's possible to take automation too far, don't overdo it. If unexpected things are happening, you need a way out quickly.
• Leverage the possibilities you have today. Then you standardize and now it's beginning to scale.
• Stay tuned for part 2 with Marc where we meet the team involved in their scrappy automation efforts and talk about what they found.

Links & Resources

• Marc Berenson on LinkedIn : https://www.linkedin.com/in/marc-berenson/

Full episode transcript

Daniel Jester:
From Creative Force, I'm Daniel Jester, and this is the eCommerce Content Creation podcast.

Daniel Jester:
We have talked a lot on this podcast about agility and scalability and automation. It can sometimes sound like a huge undertaking to pivot a large in-house studio towards automated processes. Automation doesn't have to be a huge project. Automation can begin with the tools any studio is using today. Mark Berenson joins me on the podcast for this episode to talk about some of his learnings during his time at Stanley Black and Decker and what he's been working on since,

Marc Berenson:
But the philosophy needs to change and that you need an automation mindset, and we don't get that from leadership. They're just saying, can you work faster? Can you work faster? So now it's up to us to, say, yeah, how do we start using automation? How do we get that mindset? And then it has to say, well, at the very beginning. So we're in the capture sessions, how could we use automation? How could we identify ways to make transformation with automation without all the downfall, the explosions that can happen when you go off track.

Daniel Jester:
Before we dig in with Mark, I'd like to remind our listeners of some upcoming events. First up, the Henry Stewart Photo Studio Ops Forum 2022 in New York City on May 4th. We will be recording a live episode of this podcast during the event. I'm very excited to do this. It's something that we've been talking about doing since almost day one of launching this podcast, doing a live podcast recording at one of these events. So I hope to see you there. You can take $100 off your registration fee with the discount code content pod 100. That's right. Your boy has his own discount code, content pod 100. I don't know why I'm so excited about that.

Daniel Jester:
The following day on May 5th, also in New York City is the Pixels Flow event where I'll be moderating a round table discussion at that event. It's looking like that round table's going to shape up to be something post-production centric, some good panelists lined up so far. I don't want to share anything yet because I'm not sure what is locked in and what is still up in the air, but going to be another great conversation, another great event. So those two events, back to back, in the Big Apple, New York City, hope to see you there. Now let's dig in with Mark Berenson.

Daniel Jester:
This is the eCommerce Content Creation podcast. I am your host, Daniel Jester, and joining me for this episode, Mark Berenson, formerly of Stanley Black and Decker. Mark, welcome to the show.

Marc Berenson:
Thank you so much for having me.

Daniel Jester:
It is my pleasure, Mark. Before we get into the topic, the reason that we brought you on the show today, I wanted to do something a little bit new for this podcast, but there were some big news that dropped earlier this week as of the day that we were recording this, that Pro Photo has acquired Automated Image Device Company Style Shoots. So I know that there's some of our American listeners who are familiar with Style Shoots. They're not around here in the US. I think they might be a little bit more popular in Europe, and we certainly have some listeners in Europe who are aware of Style Shoots. But Mark, for me, this was pretty big news. I wasn't expecting an announcement like this. Pro Photo really signaling I think some investment into automated workflows above and beyond lighting devices, which they are arguably the industry standard for the most commercial studios.

Marc Berenson:
Oh yeah. Pro Photo, they're the go-to's for anything strobe, anything lighting, there must be something to it. I'm not that familiar with it. I'd love to take it for a test drive. It'd be really exciting to see what comes down the road with it.

Daniel Jester:
I visioned a future and I actually talked about this with Patrick during the Pixels Flow event back in February, but the first time that I ever played with that Pro Photo, they had a pack and strobe that came out that could be controlled with software on the computer. So you had, I think it was an ethernet cable of some kind that was connecting the computer to the pack itself, and there was a little Pro Photos software that you could adjust the pack settings from the computer without having to walk over to the pack itself. It was really great for a lot of workflows where your capture station isn't close to where the pack needs to live. And I remember thinking at the time, and I talked about this with Patrick, it would be so cool to integrate that with Capture One and treat the output settings of the power pack like a next capture adjustment, so to speak.

Daniel Jester:
So treating the settings on the strobe itself just like how you have adjustments attached to individual images and Capture One because that's, Mark, as you know, that's the one thing that we can't account for. We can copy and paste adjustments from one image to another to make sure that we're consistently applying these images. We can build presets, we can do all of the stuff, but the missing part of the puzzle was what was that strobes output setting set at? And I don't know, maybe Patrick listened to me. I feel like this was probably in the works long before February, but it'll be really interesting to see where this goes, what they decide to do with Style Shoots, that both the devices and the underlying technology.

Marc Berenson:
Yeah, there's a lot there. I think it needs to be really looked at in its totality. Maybe there are users or companies out there, this could be a really good fit for, I'm not sure who, because again, I'm not as familiar with it. It's like a one stop shop and if they can perfect it and they can give people what they're looking for, then great. It helps the industry, helps us all work faster.

Daniel Jester:
Yeah. Very interesting. Very excited to see what comes out of it. But I guess that is a little bit what we're here to talk about with you today, Mark. So to set the stage for our listeners, Mark, you reached out to me on LinkedIn and I said some pretty interesting things in my LinkedIn messages about automation and agility and scalability. And we decided to set up a meeting and talk with you about some of the things that you had been thinking about and working on. We invited you to come as a guest on the show to talk about exactly that, and what you laid out for us in that pre-production meeting that we had was what struck me as a roadmap for a new or younger studio or a studio that's been around for a few years that is struggling to figure out how to automate and how to build a system that they can scale that can meet the needs, the changing needs day to day of the studio and the business itself.

Daniel Jester:
And that's what I wanted to talk to you about on this episode of the podcast. And just for our listeners. This could turn into a two part episode where we get into some of the actual specifics of some of the things that Mark and his team, it sounded, Mark, correct me if I'm wrong, like some team members from your crew at Stanley Black and Decker, and maybe some other relationships that you had really thinking about automation and how to approach automation in a way that is organic and adds value to the studio and leverages the tools that you have already.

Daniel Jester:
But I'm going to launch in. I'm talking way too much for this episode for me. So I'm going to launch in with our first question. Mark, when a studio begins to think about automation, they want to build something that is agile and scalable, where should they look first?

Marc Berenson:
So I'm not going to try to make this a lighthearted answer, but it's in the philosophy of how you work. We are the type where we get on set, shoot, send it up for clipping if that's it, or now there's some new, great automation for our ruling backgrounds, but then comes the process of retouching. But the philosophy needs to change in that you need an automation mindset and we don't get that from leadership. They're just saying, can you work faster? Can you work faster? So now it's up to us to say, yeah, how do we start using automation? How do we get that mindset? And then it has to say, well, at the very beginning. So we're in the capture sessions, how could we use automation? How could we identify ways to make transformation with automation without all the downfall, the explosions that can happen when you go off track?

Marc Berenson:
And it's really just looking at it from end to end of what each person's doing. Do they need to do that from start to finish or not? So one example is we think of retouchers, and retouchers take what we give them, it's clips, they work on it. But really instead of just clipping, why don't we think of adding on the tone curves and some basic adjustments and building those files for the retouchers so we could get them to the 70, 80% mark and then they do what they're paid to do is retouch? We don't need retouchers to add contrast to these hundreds upon hundreds of images. Can we automate that?

Daniel Jester:
Yeah. What you're describing is this idea that I've certainly used this approach in studios before as well, where it's okay, what are the things that we have to do every time? Clipping is super easy, we know that, and the industry knows that. The industry has hundreds of options on how to deal with clipping, both automated software, devices, services, individuals, all that sort of thing. But then to your point, it's okay, well, as long as we're in there doing that, we may not know exactly what the settings on this contrast curve need to be, but we know we'll need it. So why not bring a file back into our retoucher that already has the baseline layers that they need to start working?

Marc Berenson:
In addition to that, you want something that's going to work for all your images. You want something that if it's an all black item or it's all white item, or it's the mid tones grays, how do we get that agile, scalable thing? Well, so what we'll do is with the help of my team and there was a lead retoucher who worked with me, we broke down curves into three subsets. We made a preset for your shadows, a preset for mid tones and presets for highlights and that's added automatically. And each one of those was at, let's say the 50, 60% so that you could add or take away.

Marc Berenson:
But by isolating just tone curves into three different layers, we right now added something that's scalable. Meaning, because we have a bunch of black items, we just know, oh, we just unclick the black box. Or a bunch of white items, we're going to pay attention to this area. So we never went too far with anything. And we really did this heavy lifting mentality so that when we were building things, you would never say, oh, we have to take this off. This doesn't work.

Daniel Jester:
You shared this with me during our pre-production call, you shared your screen, and you showed some of these files that had these, which I thought was a really smart way to standardize the way that a file gets prepared for retouching, which is what you're saying is that you isolate the various areas of the images that you might need to apply a tone curve in a certain way or whatever the adjustment is, but leave the flexibility there. Obviously doing this, everyone's going to roll their eyes when I say this, because this is the most obvious thing ever, but obviously doing this in a way that is lossless, that we're not impacting the image in a way that can't be undone, but we can go in and just turn on and off the things that we need, make adjustments to opacity in the way that we need it.

Daniel Jester:
But what you showed me, and fortunately, doesn't do a lot of good in the medium that we're sharing this information with our listeners now it's not a visual medium it turns out. But what you showed me was really cool in the way that you showed me how you had all of these different types of images, some of which had more shadows and some of which had more shiny and reflective areas. And the layer structure was the same. The way it was all set up was the same. That's the standard. That's the 80% that we're talking about. The 20% is the retoucher going in and looking at it and saying, which of these settings looks better? What can I massage to get this across the finish line? And all of this is using tools that every studio is using already. And that's the thing that I thought was really great about it is focusing your early automation efforts on the tools and technology that you're using today, Capture One, Photoshop, whatever else it might be, seems a great way to start to build an automated process that really adds value.

Marc Berenson:
It is. And what we found is that, well, why don't we piggyback off of money and time and effort that Photoshops use? So Photoshop now has it that you can make masks based on shadows, mid tones, and highlights. So when we're making a curve for shadows, we then had the Photoshops select for the shadows for us and then we went up the scale. So Adobe invested a lot of money into that technology, we just cherry picked it.

Marc Berenson:
So part of what I'm here today, or just in general, what we were working in the studio is how do we take what's out there in existing, how do we cobble it together so that we're better today that we were yesterday, and we can continue to keep going till we hit our roadblocks, till we hit our thresholds?

Marc Berenson:
There was a team of people and we were able to just keep, well, can't you do this? Well, what if you do this way? Could you use Box files and then send them that way? We just kept playing and slowly we kept seeing more daylight and basically product on white tabletop photography got 25 to 35% faster before we were finished.

Daniel Jester:
And critical to this process, it seems to me, Mark, would be getting somebody who has a little bit of a bird's eye view of the process, who's touching on what the photographers or digi techs are doing on set in Capture One, what the retouchers are doing, can you tell me a little bit about your thoughts on this? It seems like you need to get somebody out of the weeds a little bit in starting to identify everybody's doing this one thing, but doing it a little bit differently, let's take the best pieces of all of it and create a standard.

Marc Berenson:
Yes. But when you say standard, let's really look at what standard is. Every set that shooting tabletop is using the same lighting, the same setting, same two strokes on the side, one on top, use the Plex seal light underneath it. We're all shooting in Capture One Pro, okay, well what could we do in Capture One Pro? Can we start adjusting the tone and maybe saturation or other things that would work on all images? And so once we decided to do that, we then baked into C1 Pro every camera that goes on that set automatically levels and adjustments could be set so there's never user error. So right there, we've standardized. So then when we add clipping paths, we have to do quality control in those clipping paths to make sure it a hundred percent.

Marc Berenson:
Well, the person who's doing quality control on that, doesn't have to be a senior retoucher. It could be a digi tech who's not working that day. It could be anyone in the studio. But then once that person's there and they verify that the clipping pass have integrity, why not add the tone curves and the build out that the retouch will then use? But when you've already passed off this file, it's been vetted that it's correct, it has a correct path, you've dropped out the background, you've added white, you've added tone curves, and any other adjustments that might be needed for that type of studio work.

Marc Berenson:
Standardizing means it's domino effect. You start at the very beginning and you're taking it all the way to where the expertise of your senior retoucher comes in.

Daniel Jester:
One of the things that you mentioned in there struck me as a pretty critical part of building in a process or evolving a process to include more automation, which is that QC part. Obviously all through the production process, there's all sorts of moments where you can apply QC and studios do today, but really having, like you mentioned, getting image files back from if you sent them out, maybe to get the paths done, and then you're bringing them back in, taking that extra moment of Let's look through all of them and just make sure that we're ready for the next step, as opposed to jumping into the next step.

Daniel Jester:
And then you get a third of the way into your images and you find some paths that are just completely way off the path so to speak, pun intended, definitely. It strikes me that QC process becomes an area where almost of investment of a little bit of time, because you're getting way more value from that time since you're saving so much time in other places that you do need to build back in that let's have a person take a moment and just check and make sure that our automation didn't go off the rail somewhere.

Marc Berenson:
Yeah. And the thing is even within automation, the first thing we did was the process of checking pass, we build an automated workflow. And in that case, it looked like the F keys were the way to work. So you push F1, you immediately see the integrity of the background. You push F1, shift, you're off of it. Push F2, you're dropping out your backgrounds, adding light. You push F3 and anything you want to add or is what we did, we took it a little bit further to start using scripts, to add in the tonality and maybe color adjustments and things like that.

Daniel Jester:
Right. Very, very interesting. We could have started this way, but I wanted to come back to this, but as we're talking about some of these ideas in automation, and we talk a lot on this podcast, as you know, Mark about the idea of agile and scalability. Sometimes, especially on this show, the way that we talk about it can feel like a thing that very big companies or very big studios start to think about and invest in, but let's define agile and scalable a little bit because there definitely is an application to any studio of any size who just wants to be able to work in a way that's smarter. So when we talk about agile and scalable, am I incorrect in saying, Mark, that really what we're just talking about is processes that are repeatable and that we can automate, but we're not locking ourselves into anything that we can't undo in some way?

Marc Berenson:
It's funny. Agile is everywhere. I've listened to many, many of your episodes and I see the word thrown around quite a bit and scalable, that's what eCommerce is about. I like to bring it back in this case. It's well, do it yourself, DIY approach. And within that, really what's driving this is faster delivery, that yes, it improves consistency and quality. And ultimately it's saving the company money and being more productive. That's why it has to be flexible and scalable or agile and scalable is because it needs to do those things. And it can't just do it part of the time. It has to do it on a hundred percent of the images. So it's better to know a hundred percent of the images get to a certain threshold 100% of the time, and then we work on the other percentage in fine tuning.

Marc Berenson:
But because of building in that it is agile and scalable, and that's rock solid to your foundation, it's like we're building houses. We're not renting apartments. This is a marathon, not a sprint. And so that's what agile and I think scalable are, is that a person, where we were going, which unfortunately we never got to launch fully was that we were going to implement this in the UK, in Argentina and also to one of the new studios in Taiwan. And that was where this can go. Is that because it's so defined for what works for us, but then I'd say to everyone else, it's what works for your company, your brand, that if you have three sets going all the content that's coming for those three sets, when it's going through the process, they're all going to have these foundation that's the same, and it started to capture. So if anyone needs to make any adjustments or edits a little bit later on, they already know the steps that led to it. They know exactly where to go to undo that.

Daniel Jester:
We are going to ask you to lend a bit of your experience and expertise in some of this idea, Mark, and I am going to ask one of the things, my personal experience, especially with using things like Photoshop scripts and things like that is that you have to just be aware of the weak points potentially, or the possible failure points. So in your experience, Mark, what should a studio team be aware of? What failure points when they're looking at things like automation, what are some of the pitfalls or things that maybe you think they can watch out for to avoid making some of the mistakes that maybe you or others have made in the past?

Marc Berenson:
Okay. So I made all the mistakes. I am the living mistake maker. Trust me, what happened at Stanley Black and Decker was slowly from being a generalist, I went into the product on white realm. And when I realized that it could be done faster, and it was pretty much myself and a few other associates, I just kept looking at this and saying, well, script seemed to work. Let's just build scripts and see what happens. So I build all these complicated presets. And then I worked with one of my associates to get a script going. And then I just saw, oh, they can just get overloaded. We can just get a bunch of images that look like a disaster.

Marc Berenson:
So what I would say is the failure point is complexity. So it's making small, incremental growth, and then you want to join each one of those presets or connect the automation in ways so that if something gets tripped up, if there's an issue, because everyone on the team was part of building this and knows it in and out, it's like a mechanic who knows everything under the hood, they can go in and make adjustments. So it's not like, oh, Rob made the script. He's on PTO today, and we just had a meltdown on this rush, requested as all these images, it's not working well.

Marc Berenson:
So what we did is when we build the scripts, Rob, that's a team member for us, he was able to build it so the user could pick and choose what presets get added together in his script. So his script wasn't a script as much as you can pull together and cobble all these verified presets that are guaranteed not to be too complex on top of each other. And the way that they'll come out, you get something a little bit more complex, but you used a bunch of simple steps that were additive.

Daniel Jester:
That was exactly what I was thinking in this regard too, is that I think anybody who has any amount of experience in the industry has had that experience of a Photoshop script or action that you're trying to apply to a batch of images fails at some point in the process, and it turns out there isn't a great way to know where it failed or how, and you end up finding yourself having to go through and rerun everything.

Daniel Jester:
And so I think you're right on, I think breaking some of those things down into individual building blocks, so to speak, that can be applied when they're needed and in ways that will benefit the nature of the image that they're being applied to. That's a great way to work around that. As we are wrapping up the conversation on this part of it, Mark, and I want to mention again, that we've talked to Mark about getting some of the other members of your team to dig into some of these things and talk specifically about some of these ideas that we touched on today in a future episode where we can coordinate getting everybody all together. But for this preview to that, is there anything else that you want to touch on as we wrap up the episode?

Marc Berenson:
Sure. I think little changes and a lot of little changes are big changes, especially if you're doing them time and time again. And that's what e-com is about. One season's white shirt, the next season's black pants, they don't change that much. But if you're shooting white shirts again, isn't it nice to come back to that season knowing, oh, this is how we shoot white shirts, and it's not so much from the knowledge and experience of, oh, these are my settings at work. Is that what we have as a workflow can handle these white shirts, can handle the black pants in the same time, and yet we can all move on to the more creative, the more exciting parts of our work. But there's no doubt that any studio out there can really harness small baby steps and you'll see it. It takes a while, it takes work, but you'll get there. You'll make improvements. And then you just keep getting hungrier for bigger and bigger improvements.

Daniel Jester:
Very good. Very well said, Mark. Thank you so much for coming on with us. And like I said, we are excited to have you back on to dig a little bit more. I'm sorry if it sounds to our listeners, I'm being a little cryptic. I'm not trying to be, but we wanted to break up this conversation with Mark into a little bit of the philosophy of automation and then dig into some of the specifics that him and his team worked on. So very much looking forward to having you back on to dig into that conversation, Mark, and thanks for your time today.

Marc Berenson:
Thank you, it was a pleasure. I got to tell you, we listen to the show all the time. We've learned so much from it. If I can share and just give a little bit back to so many other of your guests and companies that have shared their inner successes, their failures, really were in it together. So it's the least I could do. And thank you so much for having me.

Daniel Jester:
That's it for this episode. We will be releasing a part two of this conversation with Mark, and we'll be bringing the rest of his team to talk details on some of the tools they developed to bring automation and agility to the tools that they were already using in the studio. Many thanks to our guest, Mark Berenson, and thanks to you for listening. The show is produced by Creative Force, edited by Calvin Lands. Special thanks to Sean O'Meara. I'm your host, Daniel Jester. Until next time, my friends.