Guide: DSLR or Video camera? (work in progress)

[.spider.]

A f/1.4 lens for a tiny sensor doesn't perform as well as a f/1.4 lens for a larger sensor.  Look at their physical dimensions.  Aperture values are just ratio between focal length and aperture diameter, they do not indicate the amount of light falling on the sensor.

 

I don't know what to say but that's just wrong and photography and physics 101.

[ALwin]

I don't know what to say but that's just wrong and photography and physics 101.

 

And how long have you been working as a photographer?

 

I don't have my Nikon at home with me, however I do have my Canon G16 (1/1.7" sensor) and my Fuji X-E2 (APS-C sensor).

 

 

Fuji X-E2 settings

18mm

ISO 200

f/2.8 (Aperture Priority)

shutter speed calculated by the camera for good exposure = 1/15 second

 

Canon G16

6mm

ISO 200

f/2.8 (Aperture Priority)

shutter speed calculated by the camera for good exposure = 1/2 second

 

If I shoot the same scene again with the Canon G16, now setting to the widest aperture possible which is f/1.8 the camera calculates the exposure time to 1/5 of a second.

 

 

Link to original jpg files: http://we.tl/SiZ8lyWdbW

 

The math for the aperture is correct, however just like there are equivalent focal lengths for different sensor sizes, there are also equivalent aperture values for different sensor sizes.

[Scruffy90]

I think I mentioned a tiny bit about formats, but that mentioned only H.264 and AVCHD. But I assume you want to know in more detail about RAW, Lossless, Lossy, compression formats, etc. and how bitrates relate to video quality?

I'm familiar with the formats and bitrates. I just think that because so many cameras shoot different formats at a similar price point (I've had friends who bought the black magic cc because it shot cinema dng, then regret the workflow), that it could be beneficial in your guide.

This will be an awesome guide ones its done☺. What else were you considering adding in?

[ALwin]

I'm familiar with the formats and bitrates. I just think that because so many cameras shoot different formats at a similar price point (I've had friends who bought the black magic cc because it shot cinema dng, then regret the workflow), that it could be beneficial in your guide.

This will be an awesome guide ones its done☺. What else were you considering adding in?

 

Ok will think about it, but I know a lot of it will be over the level of people who come to this section and ask what camera to buy.

 

Usually I work with the formats that come with the cameras I choose to operate.  Cinema DNG (a.k.a Video RAW) is something I haven't work with at all yet, but I have a pretty good idea of how the workflow will be like.  And allowing for memory card storage space I always work with the highest bit rate capable.  Because I know I will be editing my videos and converting them to an output/broadcast format which will result in unnecessary bit data in the video files being discarded.  What comes strait out of the camera, almost any camera, is not ideal for publishing immediately.  It may work for connecting the camera directly to a large screen and displaying, or even viewing on a computer, but publishing to online media servers like Youtube, Vimeo, or even television broadcasting, all of them require conversion.  And conversion from one format to another always end up with bit data being discarded to some degree.

 

So far I haven't seen anyone asking about what follow focus, rails, rigs, microphones, deadcat wind covers, tripod heads, audio preamps, etc. to buy yet.  Most just ask for a camera and a basic kit lens.  And all of these are just as important, to some degree or another, as understanding the recording formats and editing workflow.  I have also yet to see anyone asking about NLE workflows.

 

I have to think about how to cover certain topics in a general guide, especially topics that may be better discussed as a form of Q&A.

 

And if you have any other specific areas you feel I should cover in a guide, let me know.

[.spider.]

there are also equivalent aperture values for different sensor sizes.

rofl ok,

1. Somebody with advanced knowledge wouldn't have chosen this scene for testing.

2. The Canon's image is much brighter.

3. ISOs are not the same on different cameras.

4. The Canon's real shutter speed is 1/2.5s.

5. The stated aperture isn't always correct.

6. Lenses are having different t-stops depending on real aperture, elements and coating. 

[ALwin]

rofl ok,

1. Somebody with advanced knowledge wouldn't have chosen this scene for testing.

2. The Canon's image is much brighter.

3. ISOs are not the same on different cameras.

4. The Canon's real shutter speed is 1/2.5s.

5. The stated aperture isn't always correct.

6. Lenses are having different t-stops depending on real aperture, elements and coating. 

 

It was the nearest scene I could get on a weekend where I am too lazy to do anything.  If the Canon's image is much brighter it should be using a faster shutter speed, which it isn't.

 

If you don't like the results of my quick test, do one yourself.

[.spider.]

If the Canon's image is much brighter it should be using a faster shutter speed, which it isn't.

 

Obviously you didn't even read my post.  

[ALwin]

Obviously you didn't even read my post.  

 

I still don't see you showing me any test you have made.  I did a test using two cameras I had on hand, and controlling only the settings I could control.  I can't control how different manufacturers design their sensors or internal image processing software.

 

And years of actual photography experience and observations have always kept me choosing my larger sensor cameras when I am working in low light.  I don't pick up small sensor cameras for low light work, even if the lenses for them are capable of wider focal length to iris diameter ratios.

 

So while on paper, all f/1.4, f/1.8 lenses should be capable of doing the same thing, in the actual world a larger sensor will always beat a smaller sensor when it comes to capturing more light.

 

Otherwise I wouldn't need to spend $3000 on a camera body and another $2500 on a lens for doing some photography of the night sky.

[.spider.]

 in the actual world a larger sensor will always beat a smaller sensor when it comes to capturing more light.

 

Not if you are using line skipping for resizing, but obviously you don't care about anything because now it starts all over again. 

[ALwin]

DSLRs are using line skipping for rescaling the image thus a bigger sensor doesn't improve low light capabilities that much.

 

By the way, there are other ways that DSLR sensors have better low light capability than most consumer level video camera sensors.  That is their ability to use higher ISO.  Most consumer video cameras have a smaller ISO range than DSLR cameras.  But that doesn't mean the quality of the image remains the same.

[ALwin]

Not if you are using line skipping for resizing, but obviously you don't care about anything because now it starts all over again. 

 

You're the first person I've heard ever to claim that DSLR video line skipping affects low light performance, because everyone else I've heard talk about line skipping talks about it affecting image quality (anti aliasing, moire) and never about low light performance.  Either you are making a statement that no one else cares about/not true or I haven't been listening to people enough.  So do you have any articles supporting your statement?

[.spider.]

You're the first person I've heard ever to claim that DSLR video line skipping affects low light performance, because everyone else I've heard talk about line skipping talks about it affecting image quality (anti aliasing, moire) and never about low light performance. 

Probably because it's common knowledge that line skipping...

can't average noise out. 

 

And scaling methods which are averaging pixels...

are also averaging noise when the source's https://www.dropbox.com/s/iinuekcwd80e3vt/noise%20source.png?dl=0 resolution is higher, what as surprise. 

[ALwin]

Probably because it's common knowledge that line skipping...

 

can't average noise out. 

 

And scaling methods which are averaging pixels...

 

are also averaging noise when the source's https://www.dropbox.com/s/iinuekcwd80e3vt/noise%20source.png?dl=0 resolution is higher, what as surprise. 

 

Ok so how does this affect your real world usage/production?

[.spider.]

Ok so how does this affect your real world usage/production?

No, I don't care about noise, I think nobody cares about noise especially at higher ISOs.

 

What a question...

[ALwin]

No, I don't care about noise, I think nobody cares about noise especially at higher ISOs.

 

What a question...

 

You make a bold statement, which covers something very technical that people seem to care little about in real world application, and now you don't care?

 

Ever since you made your statement I have been searching online to see if I can find any discussion of how line skipping and scaling processes used by DSLRs to record video affects low light performance in any way.  I found 0 articles discussing this, and I have found several discussing anti-aliasing, moire and image quality.  I don't claim to have 100% knowledge of everything, and when I discover something I don't understand or know, I make it a point to learn about it.  I do know how DSLR's with large sensors and high pixel count scale the captured image internally to produce a 1080p video, I know the negative effects such methods can have on image quality.  But I am having a hard time finding any article or discussion regarding what you are saying.  So either it's not true, or it's not on people's radar, or it doesn't matter in real world applications.

 

I mean on paper, sure the new Sony A7r II has the best camera sensor so far (according to DxO mark), but that doesn't mean I have to go replace my Nikon D4 and entire collection of lenses when I know that I can produce great images with either camera.  This is the same with your statement, something that looks negative on paper doesn't mean it will apply when being used in the real world.

 

I am comparing the capabilities of sensors which are tiny in consumer video cameras to larger ones in DSLRs when I say larger sensors provide better low light capability.  Which is a true statement.  Of course if you compare a Super 35mm video camera sensor to a full frame 35mm camera sensor, the difference is not very big.

[.spider.]

You make a bold statement, which covers something very technical that people seem to care little about in real world application, and now you don't care?

 

Ever since you made your statement I have been searching online to see if I can find any discussion of how line skipping and scaling processes used by DSLRs to record video affects low light performance in any way.  I found 0 articles discussing this, and I have found several discussing anti-aliasing, moire and image quality.  

You don't know irony?

 

 

From that, I can already conclude that the 5DmkIII reads out all its sensels, i.e., does no line skipping. However, I didn't run a resolution analysis for the 5DmkIII. However, hearing about resolution complaints for 5DmkIII video, I think they bin pixels before read out. This improves noise and aliasing performance but unlike downsampling, doesn't help the resolution.

http://blog.falklumo.com/2012/04/lumolabs-nikon-d800-video-function.html

[ALwin]

You don't know irony?

 

 

 

 

http://blog.falklumo.com/2012/04/lumolabs-nikon-d800-video-function.html

 

I did research because I thought I missed something and so that I could improve the guide... but I guess I didn't miss anything after all.

 

As for the article, obviously there would be noise as he was recording video at high ISO.  Even a still image suffers from noise when being shot at ISO 12800.

 

Also he is comparing 2 cameras with almost identical sensor sizes, albeit different pixel counts.  Your original statement was about sensor size.

 

DSLRs are using line skipping for rescaling the image thus a bigger sensor doesn't improve low light capabilities that much.

[.spider.]

I give up you don't get it. I think you don't even know how lineskipping works.

I did research because I thought I missed something and so that I could improve the guide... but I guess I didn't miss anything after all.

You are missing that lineskipping is making the sensor size irrelevant, only the dot pitch is relevant.

 

As for the article, obviously there would be noise as he was recording video at high ISO.  Even a still image suffers from noise when being shot at ISO 12800.

Irrelevant

 

Also he is comparing 2 cameras with almost identical sensor sizes, albeit different pixel counts.  Your original statement was about sensor size.

Irrelevant

Do you even know how lineskipping works? Why are you ignoring the noise images? If you don't like the results of my quick test, do one yourself.

[ALwin]

snip

 

Why don't you go back and read your first statement. Sensor size doesn't always equate to pixel count or density.  So when I say a larger sensor will tend to have better low light performance, I am comparing it to a small sensor like the ones you find in a point-n-shoot camera or camcorder, or shall I simply assume English is not your first language?

 

As for line skipping, I understand it.  But I am not going to go into detail in my guide because it's a technical topic that is beyond the level of the people who visit this forum.  If people want to learn more, they can go study where ISO comes from, how camera sensors work, etc.  They don't need to see it in a guide that is trying to teach them the differences between using a photo camera for video and a dedicated video camera.

 

You're tests do not prove or tell me what I don't already know.  And I have no need to test it when I have applied it to real world work several times. In case you are wondering, I also own and use a Nikon D800E, a 36MP camera that has a very high pixel density.  But I don't pixel peep like some people who get obsessed over every tiny detail and get stuck on technicalities.

 

This guide is not about whether someone should buy a 10MP camera or 100MP camera for video.  It's about letting people know the differences between a photo camera and a video camera.

[.spider.]

Why don't you go back and read your first statement. Sensor size doesn't always equate to pixel count or density.  So when I say a larger sensor will tend to have better low light performance, I am comparing it to a small sensor like the ones you find in a point-n-shoot camera or camcorder, or shall I simply assume English is not your first language?

 

Yes just ignore the scaling and dot pitch, it's fine.

I give up bye.

[ALwin]

Yes just ignore the scaling and dot pitch, it's fine.

I give up bye.

 

You do realize that both still cameras and video cameras with larger sensors will suffer from the same issue.  The only thing is video cameras don't tend to support very high ISO like still cameras, so that may mitigate the issue.

 

EDIT: I should say larger sensors with high megapixel count sensors.  E.g. Sony NEX VG900

[.spider.]

You do realize that both still cameras and video cameras with larger sensors will suffer from the same issue.

Nope nope nope you are driving me crazy. Hello ignore list.

[ALwin]

Nope nope nope you are driving me crazy. Hello ignore list.

 

You drove yourself crazy, if you had started with a comment about pixel density I would've understood immediately. But you started your comment about sensor size... it's like you were trying to talk about oranges but kept describing an apple.

[.spider.]

You drove yourself crazy, if you had started with a comment about pixel density I would've understood. But you started your comment about sensor size...

I highly doubt it.

[ALwin]

I highly doubt it.

 

You definitely drove yourself crazy by

1. incorrectly stating a concern regarding a very technical issue (dot pitch/pixel count is NOT sensor size)
2. a technical issue that is above the level of the average consumer
3. a technical issue that many real world professionals have found ways to work with or around it

Sensor size, pixel density, dot pitch, all of these are not black or white technical issues that determine the capabilities of a camera's sensor. Camera manufacturers have a good understanding of what they mean and what problems they can cause, so they have designed ways for the cameras they manufacture to compensate for any potential problems.  Different manufacturers apply different solutions.  Or else we would be seeing a linear relationship between higher pixel count and low light/noise performance.  Look at the new Sony A7rII, 42 megapixels and it seems to perform quite well in both still image and video recording capabilities.

 

i.e. A low megapixel camera such as the Nikon D90 should perform better than the Nikon D7200 due to the D90 having half the pixels hence larger pixel sizes and wider dot pitch. Yet in reality the D7200 performs much better than the D90.  When Nikon announced the D800/E, even before it arrived in stores, there was a lot of noise from people saying how bad it would be because Nikon packed 36MP into a FF sensor, when it arrived in stores and people started using them... they shut up.  When the D4 was released, I was disappointed with the issues regarding it's 1080p capabilities, that didn't stop me from using it for work.

 

When I entered the world of digital photography, I used to read articles about such technical issues.  Now, I could care less.

 

Do people need to know how to read MFT charts or understand what dot/pixel pitch means?  FK No!  Sure it would be some trivial knowledge they could/should have, but it's not essential to going out and take a photo or record a video.

 

As I have stated several times, cameras are just tools.  What someone creates by using that tool depends on their skill, knowledge, understanding and most important of all, their creativity.