How high do you rate MX?

[Ryan E. Walters]

This is probably a stupid question, but what do you mean, "meter" and "histogram" on these different tests? Exposing via a light meter versus exposing via the internal histogram?

The tests are very interesting, I would like to know more about them. . .

What do you think of neatvideo???

Thanks.

Stephen

Everything is explained in detail on my blog, as I did extensive testing with the MX, M, and Build 30.5.0:
http://www.ryanewalters.com/Blog/files/MXTesting_Part01.html

The blog post that relates specifically to this video is at:
http://www.ryanewalters.com/Blog/files/MXTesting_Part06.html

 

[Erich Ocean]

The image on the screen is irrelevant- one doesnt even have to set the iso on your camera at anything other than 320.

That's technically true, but one major reason to set the ISO in-camera is so that the DP can communicate how they expect the footage to be developed. It's similar to writing N+2 on your unprocessed black and white film (if you use the zone system), but better, since it's part of the clip's metadata. For that reason, I think the ISO on the camera and on your meter should always match.

 

[Andrew Wilding]

That's technically true, but one major reason to set the ISO in-camera is so that the DP can communicate how they expect the footage to be developed. It's similar to writing N+2 on your unprocessed black and white film (if you use the zone system), but better, since it's part of the clip's metadata. For that reason, I think the ISO on the camera and on your meter should always match.

I agree. Also, its nice to have an idea what the footage will look like when its pushed up a stop and a half. I was just trying to communicate to SP that it has little to do with exposing by the monitor. I use a meter for everything...

 

[Erich Ocean]

I use a meter for everything...

I think that's the best advice of all.

 

[Stephen Pruitt]

I need to come and hang out on a shoot with you guys sometime.

Okay, some of my confusion has been alleviated. When you say "metering," you mean my Sekonic. When I say "metering," I mean my in-camera meters.

So, you are setting your hand-held light meter to an ISO of 800 or 1000 and then working from that point. It suddenly all makes sense. I could not understand at all why rolling up the ISO on the RED itself would make even one iota of difference, but setting the hand-held to 800 (versus 320) certainly would lead you to under-expose by a stop or two.

Thanks.

Stephen

 

[Andrew Wilding]

I need to come and hang out on a shoot with you guys sometime.

Okay, some of my confusion has been alleviated. When you say "metering," you mean my Sekonic. When I say "metering," I mean my in-camera meters.

So, you are setting your hand-held light meter to an ISO of 800 or 1000 and then working from that point. It suddenly all makes sense. I could not understand at all why rolling up the ISO on the RED itself would make even one iota of difference, but setting the hand-held to 800 (versus 320) certainly would lead you to under-expose by a stop or two.

Thanks.

Stephen

Aha. Leave it up to semantics, eh? Words always get in the way. I guess thats why we are all image makers...
:smile5:

 

[Ryan E. Walters]

I need to come and hang out on a shoot with you guys sometime.

Okay, some of my confusion has been alleviated. When you say "metering," you mean my Sekonic. When I say "metering," I mean my in-camera meters.

So, you are setting your hand-held light meter to an ISO of 800 or 1000 and then working from that point. It suddenly all makes sense. I could not understand at all why rolling up the ISO on the RED itself would make even one iota of difference, but setting the hand-held to 800 (versus 320) certainly would lead you to under-expose by a stop or two.

Thanks.

Stephen

There are two very important reasons why changing the ISO on the camera matters:

1. (Previously mentioned) If post gets a hold of it hand does rushes at the 320, or just converts the files to ProRes (where it stays for the rest of post) all of the footage is going to look underexposed and then grading it back up to normal could degrade the quality with more noise or more banding. If it is set right when you record you KNOW you are communicating to post what your intentions are for the footage. (If the original R3D's are used to grade from then it is less of an issue as they have all the RAW data to work from, but it is still good, IMO, to communicate your intention to them away way.)

2. Job & Perception protection. While YOU may know that it is just metadata, the Director may not get that, and even MORE likely the Agency people or the Client may not get that. Then they could think that you do not know what you are doing, and that the camera system is also at fault as everything they are seeing is underexposed by 1 - 2 stops or more. Trying to explain this all on set, is not a good use of your time (IMO) - it is better to just set it so that everyone is happy and your intentions get communicated to post so that you can focus on your role as the DP.

Just my $0.02 anyway ...

 

[Rob Ruffo]

There are two very important reasons why changing the ISO on the camera matters:

1. (Previously mentioned) If post gets a hold of it hand does rushes at the 320, or just converts the files to ProRes (where it stays for the rest of post) all of the footage is going to look underexposed and then grading it back up to normal could degrade the quality with more noise or more banding. If it is set right when you record you KNOW you are communicating to post what your intentions are for the footage. (If the original R3D's are used to grade from then it is less of an issue as they have all the RAW data to work from, but it is still good, IMO, to communicate your intention to them away way.)

2. Job & Perception protection. While YOU may know that it is just metadata, the Director may not get that, and even MORE likely the Agency people or the Client may not get that. Then they could think that you do not know what you are doing, and that the camera system is also at fault as everything they are seeing is underexposed by 1 - 2 stops or more. Trying to explain this all on set, is not a good use of your time (IMO) - it is better to just set it so that everyone is happy and your intentions get communicated to post so that you can focus on your role as the DP.

Just my $0.02 anyway ...

As to "1' I guess I've been lucky and have never had to work with anybody that stupid :blink:

As to "2" just takes 5 minutes, at the beginning of the day, to show people what you can do using the controls in the Video menu with already recorded footage on set - they see, they understand. Then they leave you alone for the rest of the day - even when are ARE screwing up ;-).

 

[Ryan E. Walters]

As to "1' I guess I've been lucky and have never had to work with anybody that stupid :blink:

As to "2" just takes 5 minutes, at the beginning of the day, to show people what you can do using the controls in the Video menu with already recorded footage on set - they see, they understand. Then they leave you alone for the rest of the day - even when are ARE screwing up ;-).

To each his own. But if I can have 5 more minutes to light my set, I'll take it.

 

[Sanjin Jukic]

I can rate MX to the maximum of 6400 ISO

I can rate MX to the maximum of 6400 ISO

I can rate RED1-MX to the maximum of 6400 ISO using together Neat Video, The Foundry's Furnace Core De-Noise and FCP sharpen plug ins.

Still rendering test examples.

More about that coming soon here.

Stay tuned.

 

[Stephen Pruitt]

I think Ryan's post about whys is probably a good one. In my own case, since I'm the co-writer/producer/director and I don't have private clients, it doesn't matter to me at all. BUT, I can sure see it otherwise.

Thanks again for the words of wisdom.

Stephen

 

[Tim Sutherland]

In my experience it almost never benefits to leave the camera at 320 if you are going to 'develop' it otherwise. Hair/makeup/etc all need to see a decent looking image, and I don't want people getting the idea the idea that the image has to be 'fixed' after shooting, there is nothing wrong with it.

After working with MX for almost every day since middle of April, I can tell pretty well what detail is held in the highlights or not even from the SDI tap at ISO 800/1000. I also think there's no reason to change the iso in post from what is metered/used on set except for when you need more range for a particular situation.

Changing things from metadata in post as a rule of thumb doesn't make much sense to me, because it adds a lot of possibility for confusion. If post knows that the usual operation is to keep the ISO from the metadata, then your only changes to that are the exceptions. If you do it the other way around, no matter how many notes you make, post will probably make a few mistakes from time to time, which can cause headaches.

I will say, however, that I am a fan of using a downconverter and SD playback for commercials, because the agency and client are less picky because they feel more like its film and that that is the worst image they will ever see, which is usually true.

Also my $0.02.

Tim

 

[Gunleik Groven]

Ryan. Have read your blog. Thanks!

Very interesting.

I am about to do my first serious meeting with the MX over the next few days, your tests seems encouraging to say the least. I still have a couple of questions...

Note to anybody asking: My experience is from M shooting and post.

I really have found that the ISO rating for all practical measures is meaningless. Not unusefull, but meaningless as a way to use the way you are suggesting.
No offense intended!

For shooting anything from 320 upwards (if you really need to use gain to see what you're doing) is good.

For post anything from 250 downwards to 80 (using the exposure parameter) are good startingpoints. And oddly enough for post: the higher the iso on set, the lower the exposure settings for post.

Slightly exaggerating here, but just to make a point of how I have come to think of the development of R1 - M images...

Of course that has to be compensated - for development - to get a gradeable image, and usually I would do that with a combination of Brightness/contrast/curve operations (which are lift/gamma-oriented) rather than exposure/iso - which are gain (and thus noise-generating) operations. All going to Redlog pre-grade.
And the trickier the image is, the closer I will put the WB to 5600 pre grade...

And with overexposed images I use to check them in linear-light to see what is possible to recover.

OK.
Admitted.
These are dire generalisations. As of 30.5 (and even better @ 30.6...), most clips can be batched for post without much further ado.

What I often have to do, is to deliver images suitable for grading/compositing/keying, thus not "good looking" images as such, but images that are good for the treatment they are about to get.

So what does this boil down to?

A couple of questions....

-

Except from the lower noise/higher sensitivity in the dark areas, do you think the MX conceptually performs different from the M for online work. If so, in which way?

I would very much like to have a look @ some of your raw files for the underexposed MX and M shots.

That is NOT to say that I in any way doubt your conclusions, just that it would informative for me to compare my results with yours. This is NOT for making my results public. I just would like to see them.

I will be shooting my own tests with MX and that other cam over the next few days. So this is just to gather additional knowledge.

I DO understand that my approach seems to differ from yours, though I do not say that I disagree with your coclusions... -

But again:
Thanks again for your thorough analysis.

Cheers!
Gunleik

 

[Sanjin Jukic]

As I promised before my MX ISO rating test is now finished.

Here we go:

I mostly used to rate R1-MX @ ISO 800 that has always pretty clean image.

Second to that is ISO 1600 and use when I need it.

Extreme solution starts @ ISO 4000 that has still very good image.

Maximim ISO @ 6400 is also very usable and I recommend this combo plug in workflow for getting

a clean image out of really critical from the noise defected low and available light shots.

All you can see very good at the footage that has a link below.

Keep in mind that CC or color grading is not done on those shots,

that means if higher ISO make shifting colors you should start to fix it immediately in RCX and later on grading suite of your choice.

Shot handheld on R1-MX with Leica Apo-Summicron-M 75mm f/2.0 wide open @ f/2.0, ISO 800, 25fps, 1/50 shutter, available light.

Shot handheld on R1-MX with Leica Apo-Summicron-M 75mm f/2.0 wide open @ f/2.0, ISO 1600, 25fps, 1/50 shutter, available light.

Shot handheld on R1-MX with Leica Apo-Summicron-M 75mm f/2.0 wide open @ f/2.0, ISO 4000, 25fps, 1/50 shutter, available light.

Shot handheld on R1-MX with Leica Apo-Summicron-M 75mm f/2.0 wide open @ f/2.0, ISO 6400, 25fps, 1/50 shutter, available light.

Download 1080p test footage @ 2.40:1 widescreen 1920x800, h264, 74.8 MB >>>

 

[Tim Sutherland]

In my experience, setting you meter to the same ISO as the camera usually places skintones around 50% on the SDI waveform. We usually stop down at least a half from the meter reading in daylight, up to stop down from the reading in high congrats. For tumgsten, the meter reading is close, we might stop down up to about a half for certain situations.

Hope that helps. I realize this is quite different than mysterium, but yes, MX is that good. Best thing to do is do your own tests, as you have mentioned.

Tim

 

[Ryan E. Walters]

Ryan. Have read your blog. Thanks!

No problem, always glad to help add to the body of knowledge and experience.

Note to anybody asking: My experience is from M shooting and post.

The M & MX are VERY different chips and should be treated differently, so testing is key to figure out how you want to proceed for your own working methodology.

I really have found that the ISO rating for all practical measures is meaningless. Not unusefull, but meaningless as a way to use the way you are suggesting.
No offense intended!

No offense taken- if the way you work, works well for you, great. My methodology has allowed me to produce consistent results in a timely fashion resulting in images that I and my clients are very happy with. If your way works for you, no problem at all.

For shooting anything from 320 upwards (if you really need to use gain to see what you're doing) is good.

For post anything from 250 downwards to 80 (using the exposure parameter) are good startingpoints. And oddly enough for post: the higher the iso on set, the lower the exposure settings for post.

Slightly exaggerating here, but just to make a point of how I have come to think of the development of R1 - M images...

Of course that has to be compensated - for development - to get a gradeable image, and usually I would do that with a combination of Brightness/contrast/curve operations (which are lift/gamma-oriented) rather than exposure/iso - which are gain (and thus noise-generating) operations. All going to Redlog pre-grade.
And the trickier the image is, the closer I will put the WB to 5600 pre grade...

And with overexposed images I use to check them in linear-light to see what is possible to recover.

The M & MX are really different chips- with the M I would NEVER push or pull the chip past ISO 200 - 320. I found that keeping it at ISO 320 and lighting to that, or ISO 200 if I wanted an ultra clean image was the way to get the best results out of the chip. I am very noise intolerant, so I was always more conservative with the M. This is proving to not be the same with the MX. I can push the MX a lot more then the M before the noise gets unacceptable to me. The "problem" I see with the MX and ISO is that at higher settings the highlights seem to get more blocked up - especially in a low contrast scene. (See the Re-Graded footage of the MX in my test: http://www.ryanewalters.com/Blog/files/MXTesting_Part06.html ) If post just converted the footage via the metadata of ISO 800 for that scene a lot of the detail in the sky would have been lost, looking video-ey.)

Except from the lower noise/higher sensitivity in the dark areas, do you think the MX conceptually performs different from the M for online work. If so, in which way?

Yes, they are very different.

With the M I would say shoot at ISO 320, and keep it there- NEVER change the ISO on the camera. If you change the ISO anywhere, it should be on your meter, so that you are lighting to an ISO of 200 or 250.

With the MX, I would suggest shooting at ISO 320 for low contrast scenes for the example I references above, and then shooting at ISO 800 for high contrast, and if you need to be shooting in low light, you can push the sensor even high if you like.

Basically, with the MX I would suggest populating the sensor as much as possible and with the mountain in the middle of the histogram. Use whatever ISO is best suited to get you to that point for that scene.

I would very much like to have a look @ some of your raw files for the underexposed MX and M shots.

That is NOT to say that I in any way doubt your conclusions, just that it would informative for me to compare my results with yours. This is NOT for making my results public. I just would like to see them.

Doubt my conclusion all you want - I can take it. I've been wrong before. Contact me off line (PM or email) and we can try and figure out a wat to get you some of the files. The biggest problem I have it figuring out a way to move that much data over the net, as the files are not small at all ...

I will be shooting my own tests with MX and that other cam over the next few days. So this is just to gather additional knowledge.

I DO understand that my approach seems to differ from yours, though I do not say that I disagree with your coclusions... -

But again:
Thanks again for your thorough analysis.

Cheers!
Gunleik

No problem at all. I'd be more then happy to talk about your approach and why it works best for you- I'm sure there are some things in the way you do things that I could learn / benefit from even if in the end I continue to use my methodology.

 

[Sanjin Jukic]

I just opened the new tread about getting MX high ISO @ 6400

LINK>>>