F3 and Epic low-light/dynamic-range, how do they compare? at DVinfo.net

Mark Kenfield · November 6th, 2011, 05:53 AM

Hey guys,

I'm tossing up between the idea of either an F3 (not SLOGged) or a Scarlet recording out to Pro Res, and I'm wondering if anyone's had any experience with both the F3 and Epic and can comment on how the low-light sensitivity and dynamic range of the two compare?

It's practically impossible to get any objective perspective from the Red crowd, so I'm wondering if you guys can help me.

Cheers

Brian Drysdale · November 6th, 2011, 07:26 AM

Not sure you'll get much accurate stuff regarding this on the new Scarlet, I don't think anyone outside RED has even seen pictures from the camera yet. I'd assume the quoted dynamic range figure is from the RAW recording, rather than a HD SDI output.

In making a decision, delivery of the Scarlet may be another issue, it'll really depend when you want to have the camera.

Mark Kenfield · November 6th, 2011, 07:43 AM

Cheers Brian. From everything I've heard so far, Scarlet IQ should supposedly be pretty much just a 4k version of Epic's output.

Interesting point about the RAW vs. HD-SDI dynamic range, I hadn't really thought of that aspect, I wonder how much of a difference it might make?

I'm popping into the Atomos offices next week for a demo of how their Samurai recorder performs on the Epic, so I'll see what I can learn there.

Here in Oz, the SLOGged F3 is out of my price range, so if the dynamic range from the Epic sensor (in Scarlet) can give me a level of flexibility that the standard F3 can't match, that could be a considerable boon.

Brian Drysdale · November 6th, 2011, 11:37 AM

Someone elsewhere has queried if the Scarlet does a proper debayer when it gives an HD SDI output. Just something to be checked, it mightn't be an issue, but they mightn't do the highest quality conversation in camera when RED plan for a RAW workflow.

Duke Marsh · November 10th, 2011, 09:02 PM

Might be the same person as on the 'other' forum, but in case its not:

Philip Bloom did some simultaneous shots between the F3, Epic and AF100 around a campfire at night while shooting a documentary. There's more than one blog about it. Philip does have sponsers, but he discloses them and neither Sony, nor Red, nor Panasonic are sponsers.

Shooting a documentary with the RED Epic, AF100 and F3 | Philip Bloom

"I captured this with the F3 and the AF100. This was just lit with the fire so very dark. The F3 did the best here easily. The AF100 did OK with the Abel Cine low noise profile and the amazing Voigtlander 25mm F0.95. The Epic struggled to see much to be honest. It needed more light. So I just used it for some off speed pretty shots of the fire which looked lovely!"

When asked "Why did you choose not to use the Epic on the fire sequence?" "The low light on the Epic was not up to it. Far too noisy."

It makes sense. Light sensitivity is partially governed by the size of the photo sites. The Epic and the Scarlet have 5k sensors, the F3 is approx. 2k - 3k wide. The Red cameras have tiny photo sites to fit many more sites into the approx. same size S35 sensor.

Also Philip notes "To convert a 1 minute clip (shot at 5K 2.4:1 with 12:1 compression) to 1080p on a 2.53GHz Intel Core 2 Duo, it took 40 minutes!" He much preferred using the F3 over the Epic, and the Scarlet has many more compromises than the Epic.

That's the only non-Red Fanboi review I could find.

Alister Chapman · November 11th, 2011, 03:33 AM

This is a reoccurring theme with the Red Cameras, or more specifically the Red Workflow. The Red MX/Epic etc cameras are great cameras that can produce very good images, I don't think anyone will dispute that. But the workflow is very different to what many video users are used to. It's not really instant. Red has always aimed fair and square at film, so Red is trying to squeeze every possible grain of performance out of the system (not saying that doesn't apply to all cameras). So by recording the sensors output with as little processing as possible you give the user the maximum possible manipulation possibilities with that output in post. But this then presents a very different workflow to more conventional cameras that do the majority of the image processing in camera. These are two different ways to work and you need to consider this when choosing your camera. For a film shoot the Red workflow is normally perfectly acceptable as footage isn't normally edited immediately so there is time to do the processing. Red even have their Red Rocket computer card to speed things up dramatically if you are doing a lot of Red work. However if you need a fast turn around this workflow might not be suitable.

No one knows yet how good the HDSDi out of Scarlet will be. On Epic AFAIK it's still limited to 720P with an upgrade to 1080P scheduled for a future firmware update, so recording to an external device may not be possible for a while yet.

Convergent Design did some low light test with a Red One MX compared to an F3 and the difference was quite striking with the F3 proving to be a lot cleaner. As has been said the penalty for higher resolution on the same size sensor tends to be a reduction in sensitivity, so you need to choose, do I need resolution or do I need low light performance?

Ed David · November 11th, 2011, 11:37 AM

Thank you guys for posting this honest feedback. I also don't know how exactly the image quality will be out of the HD SDI port due to the fact that the Red processes the images in post - so you may have an incomplete image coming out. It's important to be objective to all issues and I appreciate that greatly Alister.

Duke Marsh · November 12th, 2011, 07:48 AM

Has anyone had practical experience with xHDR in any of the Red cameras yet, like the Epic?

My understanding is that it creates HDR the same way a photo camera does, namely by taking two non-simultaneous pictures (one under exposed and one over exposed) then combines them.

What I'm wondering is how much ghosting will be a problem in practical terms when there is motion. I would expect it to be more of an issue with high resolution sensors. Of course it won't show on a talking head, but you don't usually need HDR on a talking head.

The claimed DR would be phenomenal, but if you can't use it with motion its a bit gimmicky.

Alister Chapman · November 12th, 2011, 09:41 AM

I've not used it, but I have seen the results. Any fast motion will cause some ghosting, pans etc appear to blur a little. If you use a high shutter speed the blur is reduced somewhat. It's not terrible but can produce some slightly strange artefacts around high contrast moving objects where they can have bright or dark edges.

If you really need HDR you can always use a pair of cameras on a beam splitter 3D rig to capture the high and low ranges in perfect sync. I'm already experimenting with a pair of S-Log F3's on one of my Hurricane Rigs. It's bulky, but the results are promising.

Duke Marsh · November 12th, 2011, 09:56 AM

The problem with using a high speed shutter is that you lose more light and can produce jerky motion. Personally I'm greatly looking forward to electronic indexing. Ned is trying a beta version and apparently it does what its rumored to do. We all, in certain situations, can use an extra stop of DR on top of the extra stop S-Log gives us.

It sounds like the choice is going to be; If you want/need dynamic range, low light sensitivity and fast turn around times you pick the F3.

If you want/need the highest resolution you pick Epic.

Don Miller · November 12th, 2011, 10:25 AM

Quote:

Originally Posted by Duke Marsh

Also Philip notes "To convert a 1 minute clip (shot at 5K 2.4:1 with 12:1 compression) to 1080p on a 2.53GHz Intel Core 2 Duo, it took 40 minutes!" He much preferred using the F3 over the Epic, and the Scarlet has many more compromises than the Epic.

That's the only non-Red Fanboi review I could find.

To be fair I believe Philip has ordered a Scarlet.

Not picking on Bloom specifically, but I'm sure all of these guys who have become public experts use all the larger sensor cameras to remain experts and "in the game". Would they be using the whole range of hot products if the camera was just a tool for a job? I doubt it.

I expect both Sony and Panasonic experience some lack of attention for their new products simply because they are established and not as exciting as the new new thing. For example the low light capability of the F3 compared to Epic. If the situation was reversed the 'word' would be about old Sony sensor tech. Yet Red lagging behind the Japanese on CMOS sensor tech. goes almost unnoticed.

Brian Drysdale · November 12th, 2011, 02:34 PM

Following the comments regarding low light shooting here, the Epic seems to lag behind both the F3 and the C300 in the low light stakes.

The RC #100 C300 with Dir. Laforet & RED ScarletX | fxguide

Very much a matter of selecting the camera that's right for the production you're shooting or for the way you like to work. They've all got strengths and weaknesses.

Duke Marsh · November 12th, 2011, 05:59 PM

"Very much a matter of selecting the camera that's right for the production you're shooting or for the way you like to work. They've all got strengths and weaknesses."

Very true. I don't think the Red Fanbois are doing anyone, not even Red, any favors by claiming the Scarlet (or even the Epic) is killing the competition. They each have their place and stories aren't told with resolution.

Duke Marsh · November 14th, 2011, 08:48 AM

Quote:

Originally Posted by Mark Kenfield

Interesting point about the RAW vs. HD-SDI dynamic range, I hadn't really thought of that aspect, I wonder how much of a difference it might make?

I've been thinking about this. The specs for HD-SDI won't handle 4.4.4 so you won't get 4.4.4 out of the Scarlet or the Epic through the HD-SDI port.

Same with the F3, but at least there you have the 3G ports which will output 4.4.4, plus with v1.3 firmware will allow you to split it for a LUT and non-LUT signal. Run one to your 4.2.2 recorder and one to your monitor (especially your client monitor.)

Betsy Moore · June 19th, 2012, 05:03 AM

Ok, this is a bizarre question and I'm not sure if this is the right forum but since the conversation is in the right ballpark...

... I haven't made a movie in a few years and the last time I did it was in 2005 Sony Fx1. Well now I have the opportunity to use an epic camera for only a token fee but I want to do the project with as few lights as possible so tonight I whipped out my old Sony FX1 and it would just barely do the job under the lighting conditions I prefer. Eo any of the Sonyphiles out there have a rough feeling for which camera is worse in low light?--my seven year old FX1 or a Red Epic X (with a 24-105mm Cannon lens)?

If any of ya'll have any feel for this I sure appreciate it:)

November 6th, 2011, 07:26 AM	#2
Brian Drysdale Inner Circle Join Date: Feb 2006 Location: Belfast, UK Posts: 6,152	Re: F3 and Epic low-light/dynamic-range, how do they compare? Not sure you'll get much accurate stuff regarding this on the new Scarlet, I don't think anyone outside RED has even seen pictures from the camera yet. I'd assume the quoted dynamic range figure is from the RAW recording, rather than a HD SDI output. In making a decision, delivery of the Scarlet may be another issue, it'll really depend when you want to have the camera.

November 6th, 2011, 07:43 AM	#3
Mark Kenfield Major Player Join Date: Apr 2007 Location: Melbourne, Australia Posts: 470	Re: F3 and Epic low-light/dynamic-range, how do they compare? Cheers Brian. From everything I've heard so far, Scarlet IQ should supposedly be pretty much just a 4k version of Epic's output. Interesting point about the RAW vs. HD-SDI dynamic range, I hadn't really thought of that aspect, I wonder how much of a difference it might make? I'm popping into the Atomos offices next week for a demo of how their Samurai recorder performs on the Epic, so I'll see what I can learn there. Here in Oz, the SLOGged F3 is out of my price range, so if the dynamic range from the Epic sensor (in Scarlet) can give me a level of flexibility that the standard F3 can't match, that could be a considerable boon.

November 6th, 2011, 11:37 AM	#4
Brian Drysdale Inner Circle Join Date: Feb 2006 Location: Belfast, UK Posts: 6,152	Re: F3 and Epic low-light/dynamic-range, how do they compare? Someone elsewhere has queried if the Scarlet does a proper debayer when it gives an HD SDI output. Just something to be checked, it mightn't be an issue, but they mightn't do the highest quality conversation in camera when RED plan for a RAW workflow.

November 10th, 2011, 09:02 PM	#5
Duke Marsh Regular Crew Join Date: Oct 2011 Location: OC, CA Posts: 65	Re: F3 and Epic low-light/dynamic-range, how do they compare? Might be the same person as on the 'other' forum, but in case its not: Philip Bloom did some simultaneous shots between the F3, Epic and AF100 around a campfire at night while shooting a documentary. There's more than one blog about it. Philip does have sponsers, but he discloses them and neither Sony, nor Red, nor Panasonic are sponsers. Shooting a documentary with the RED Epic, AF100 and F3 \| Philip Bloom "I captured this with the F3 and the AF100. This was just lit with the fire so very dark. The F3 did the best here easily. The AF100 did OK with the Abel Cine low noise profile and the amazing Voigtlander 25mm F0.95. The Epic struggled to see much to be honest. It needed more light. So I just used it for some off speed pretty shots of the fire which looked lovely!" When asked "Why did you choose not to use the Epic on the fire sequence?" "The low light on the Epic was not up to it. Far too noisy." It makes sense. Light sensitivity is partially governed by the size of the photo sites. The Epic and the Scarlet have 5k sensors, the F3 is approx. 2k - 3k wide. The Red cameras have tiny photo sites to fit many more sites into the approx. same size S35 sensor. Also Philip notes "To convert a 1 minute clip (shot at 5K 2.4:1 with 12:1 compression) to 1080p on a 2.53GHz Intel Core 2 Duo, it took 40 minutes!" He much preferred using the F3 over the Epic, and the Scarlet has many more compromises than the Epic. That's the only non-Red Fanboi review I could find.

November 11th, 2011, 03:33 AM	#6
Alister Chapman Inner Circle Join Date: Dec 2005 Location: Bracknell, Berkshire, UK Posts: 4,957	Re: F3 and Epic low-light/dynamic-range, how do they compare? This is a reoccurring theme with the Red Cameras, or more specifically the Red Workflow. The Red MX/Epic etc cameras are great cameras that can produce very good images, I don't think anyone will dispute that. But the workflow is very different to what many video users are used to. It's not really instant. Red has always aimed fair and square at film, so Red is trying to squeeze every possible grain of performance out of the system (not saying that doesn't apply to all cameras). So by recording the sensors output with as little processing as possible you give the user the maximum possible manipulation possibilities with that output in post. But this then presents a very different workflow to more conventional cameras that do the majority of the image processing in camera. These are two different ways to work and you need to consider this when choosing your camera. For a film shoot the Red workflow is normally perfectly acceptable as footage isn't normally edited immediately so there is time to do the processing. Red even have their Red Rocket computer card to speed things up dramatically if you are doing a lot of Red work. However if you need a fast turn around this workflow might not be suitable. No one knows yet how good the HDSDi out of Scarlet will be. On Epic AFAIK it's still limited to 720P with an upgrade to 1080P scheduled for a future firmware update, so recording to an external device may not be possible for a while yet. Convergent Design did some low light test with a Red One MX compared to an F3 and the difference was quite striking with the F3 proving to be a lot cleaner. As has been said the penalty for higher resolution on the same size sensor tends to be a reduction in sensitivity, so you need to choose, do I need resolution or do I need low light performance? __________________ Alister Chapman, Film-Maker/Stormchaser http://www.xdcam-user.com/alisters-blog/ My XDCAM site and blog. http://www.hurricane-rig.com

November 6th, 2011, 05:53 AM	#1
Mark Kenfield Major Player Join Date: Apr 2007 Location: Melbourne, Australia Posts: 470	F3 and Epic low-light/dynamic-range, how do they compare? Hey guys, I'm tossing up between the idea of either an F3 (not SLOGged) or a Scarlet recording out to Pro Res, and I'm wondering if anyone's had any experience with both the F3 and Epic and can comment on how the low-light sensitivity and dynamic range of the two compare? It's practically impossible to get any objective perspective from the Red crowd, so I'm wondering if you guys can help me. Cheers

November 11th, 2011, 11:37 AM	#7
Ed David Regular Crew Join Date: Jan 2008 Location: Brooklyn, NY Posts: 119	Re: F3 and Epic low-light/dynamic-range, how do they compare? Thank you guys for posting this honest feedback. I also don't know how exactly the image quality will be out of the HD SDI port due to the fact that the Red processes the images in post - so you may have an incomplete image coming out. It's important to be objective to all issues and I appreciate that greatly Alister. __________________ ____ www.kittyguerrilla.com

November 12th, 2011, 07:48 AM	#8
Duke Marsh Regular Crew Join Date: Oct 2011 Location: OC, CA Posts: 65	Re: F3 and Epic low-light/dynamic-range, how do they compare? Has anyone had practical experience with xHDR in any of the Red cameras yet, like the Epic? My understanding is that it creates HDR the same way a photo camera does, namely by taking two non-simultaneous pictures (one under exposed and one over exposed) then combines them. What I'm wondering is how much ghosting will be a problem in practical terms when there is motion. I would expect it to be more of an issue with high resolution sensors. Of course it won't show on a talking head, but you don't usually need HDR on a talking head. The claimed DR would be phenomenal, but if you can't use it with motion its a bit gimmicky.

November 12th, 2011, 09:41 AM	#9
Alister Chapman Inner Circle Join Date: Dec 2005 Location: Bracknell, Berkshire, UK Posts: 4,957	Re: F3 and Epic low-light/dynamic-range, how do they compare? I've not used it, but I have seen the results. Any fast motion will cause some ghosting, pans etc appear to blur a little. If you use a high shutter speed the blur is reduced somewhat. It's not terrible but can produce some slightly strange artefacts around high contrast moving objects where they can have bright or dark edges. If you really need HDR you can always use a pair of cameras on a beam splitter 3D rig to capture the high and low ranges in perfect sync. I'm already experimenting with a pair of S-Log F3's on one of my Hurricane Rigs. It's bulky, but the results are promising. __________________ Alister Chapman, Film-Maker/Stormchaser http://www.xdcam-user.com/alisters-blog/ My XDCAM site and blog. http://www.hurricane-rig.com

November 12th, 2011, 09:56 AM	#10
Duke Marsh Regular Crew Join Date: Oct 2011 Location: OC, CA Posts: 65	Re: F3 and Epic low-light/dynamic-range, how do they compare? The problem with using a high speed shutter is that you lose more light and can produce jerky motion. Personally I'm greatly looking forward to electronic indexing. Ned is trying a beta version and apparently it does what its rumored to do. We all, in certain situations, can use an extra stop of DR on top of the extra stop S-Log gives us. It sounds like the choice is going to be; If you want/need dynamic range, low light sensitivity and fast turn around times you pick the F3. If you want/need the highest resolution you pick Epic.

November 12th, 2011, 02:34 PM	#12
Brian Drysdale Inner Circle Join Date: Feb 2006 Location: Belfast, UK Posts: 6,152	Re: F3 and Epic low-light/dynamic-range, how do they compare? Following the comments regarding low light shooting here, the Epic seems to lag behind both the F3 and the C300 in the low light stakes. The RC #100 C300 with Dir. Laforet & RED ScarletX \| fxguide Very much a matter of selecting the camera that's right for the production you're shooting or for the way you like to work. They've all got strengths and weaknesses.

November 12th, 2011, 05:59 PM	#13
Duke Marsh Regular Crew Join Date: Oct 2011 Location: OC, CA Posts: 65	Re: F3 and Epic low-light/dynamic-range, how do they compare? "Very much a matter of selecting the camera that's right for the production you're shooting or for the way you like to work. They've all got strengths and weaknesses." Very true. I don't think the Red Fanbois are doing anyone, not even Red, any favors by claiming the Scarlet (or even the Epic) is killing the competition. They each have their place and stories aren't told with resolution.

June 19th, 2012, 05:03 AM	#15
Betsy Moore Major Player Join Date: Mar 2003 Location: Los Angeles Posts: 444	Re: F3 and Epic low-light/dynamic-range, how do they compare? Ok, this is a bizarre question and I'm not sure if this is the right forum but since the conversation is in the right ballpark... ... I haven't made a movie in a few years and the last time I did it was in 2005 Sony Fx1. Well now I have the opportunity to use an epic camera for only a token fee but I want to do the project with as few lights as possible so tonight I whipped out my old Sony FX1 and it would just barely do the job under the lighting conditions I prefer. Eo any of the Sonyphiles out there have a rough feeling for which camera is worse in low light?--my seven year old FX1 or a Red Epic X (with a 24-105mm Cannon lens)? If any of ya'll have any feel for this I sure appreciate it:)