Subscriber Discussion

Interested In An IPVM Bandwidth And Storage Calculator?

Avatar
Sean Patton
Jul 26, 2018

IPVM gets a few questions every month about where to find storage and bandwidth calculations in the Camera Calculator.

Our general concern is that bandwidth and disk space calculations are hard to accurately and consistently produce given the numerous factors involved (model variations, scene complexity, smart codecs, compression levels, to name a few). Indeed, bandwidth and disk space calculators tend to be very imprecise. (Related, see Calculating Video Surveillance Storage / Bandwidth).

However, we do ongoing bandwidth testing of numerous cameras, so we have experience and information to build off of.

We are looking for feedback and interest levels from users regarding IPVM developing our own Bandwidth and Storage Calculator, which could be standalone or used in conjunction with the Camera Calculator. We would aim to make it more accurate and better ground it in real factors that drive bandwidth and storage calcuation.

UM
Undisclosed Manufacturer #1
Jul 27, 2018

Sean, one item that will always happen when an estimator is used is that its results will be compared to another estimator.  One of the choices to be made is to decide if you want to present a 'CYA' bit rate, which will cause the result to demand more storage and other resources, or a middle of the road one which will tend to give a better answer concerning the points just mentioned. 

SD
Shannon Davis
Jul 27, 2018
IPVMU Certified

How about adding amperage calculations based upon cameras used in the camera calculator.

(1)
Avatar
Mark Jones
Jul 27, 2018

To be honest, there are quite a few out there.  I am not sure it is worth your while.  There is no harm in it.

JH
John Honovich
Jul 27, 2018
IPVM

Mark, we are well aware that there are many out there. 

The issue is that they are highly inaccurate. The question, for us, is can we make one that is more accurate and how much do people care for one that is?

Avatar
Mark Jones
Jul 27, 2018

Speaking only for me, I would care a great deal.  The few that I do rely on (other than my own) are really pretty accurate.  I am a member and enjoy many of your features.  Put this one on the list, but it would not be a deal breaker.

(2)
JH
John Honovich
Jul 27, 2018
IPVM

Mark, which ones do you use that are pretty accurate?

Avatar
Mark Jones
Jul 27, 2018

March Networks has a good one, and so does Open Eye.  

JH
John Honovich
Jul 27, 2018
IPVM

Mark, thanks. I just used the March Networks one for the first time.

 

For third party cameras, one literally needs to enter in a bitrate:

I am guessing then you are using it for March's own cameras or you enter in a bitrate yourself? For sure, entering in a bitrate yourself will be most accurate but that assumes the user already knows it.

(2)
(1)
Avatar
Anton Miller
Jul 27, 2018
Shaked Projects

Probably yes, along with couple more, to get average result.

SC
Sean Chang
Jul 27, 2018
Rasilient Systems

As a server and storage vendor, we are absolutely interested in the bandwidth calculator.

(1)
UM
Undisclosed Manufacturer #2
Jul 27, 2018

This is a great topic! So misunderstood when it comes to the details and so difficult to discern differences from one calculator to another.

If you can, in scientific method, do a comparison between the numerous calculators out there and their comparative results, it would be a lot of work but exceptionally informational.

We out here in the video design world don't really need another calculator. We need to know why one calculator shows significantly different results from a different calculator based on "supposedly" similar or identical parameters. That is a fairly monumental task given the occlusion of raw deterministic data on a video stream.

There are knowns such as recording rate and recording resolution. 

There are so many variables to precisely verify/compare such as "scene complexity," "actual smart codec efficiency," "estimated motion hours per day, "I frame interval setting," and others.

Really tough task but you guys are smart enough and have the resources to prepare a scientific or nearly scientific comparison of some of the primary calculators out there.

Go for it. What would be exceptional if you decide to do your own calculator is to provide results from other calculators side by side with same parameters.

Yes, monumental request but hey, might as well ask! I suspect legal will have a problem with this as well. Brown water is clear to some and black to others.

(3)
(1)
UM
Undisclosed Manufacturer #3
Jul 29, 2018

One, there are a lot of calculators, yes, but I would prefer a vendor agnostic one and one from someone who doesn't sell stuff. Sometimes I need to refer people, potential clients, to a bandwidth calculator. But I don't deal in March Networks, as one example from Mark Jones, so I don't want to send clients to a product page I don't deal in. And I don't want clients going to "Joe's online CCTV shop", either.

Second, you're worried about people people putting too faith much into the "accuracy" of a storage calculator if you did one. But isn't it the same concern if someone puts too much faith into the accuracy of the example pictures you display for PPF in the Camera Calculator?

A lot of variables can affect the clarity in the above sample picture; quality of lens, air density, light levels, light direction, quality of the imager, codec, compression level, etc.

So to me, if you're willing to display the possible results of one calculation that could have wide variances in the real world, there shouldn't be a problem with another calculation that could have wide variances.

I will say the same thing as Mark says, it would be a nice like to have, but it wouldn't be a deal breaker if you declined to do one.

 

(1)
JH
John Honovich
Jul 29, 2018
IPVM

But isn't it the same concern if someone puts too much faith into the accuracy of the example pictures you display for PPF in the Camera Calculator?

The problem with storage calculation is that its a specific number. If you say it is going to be 10TB and it turns out you need 20TB, the user will be upset. Related, this is why storage calculators tend to heavily pad their numbers, generally using worst case scenarios.

(1)
(2)
UM
Undisclosed Manufacturer #1
Jul 30, 2018

John is correct in that many manufacturers pad the results a lot which is both a bad and good tactic because you can 'guarantee' your amount of storage,  however, you will be using a higher bit rate which will push the system capability into the higher class machines.

At one time, our calculator tended to user higher bit rates, and sales was beating us up for suggesting too many systems.   We revamped our algorithm and it now tends to be more middle of the road and closer to reality.

The funny thing now is that from time to time we will get a complaint that it is too low when it gets compared to one of the others that has the 'padded' result.

With any estimator, it comes down to the bit rate.   All the users of these tools should verify that the bit rate they see matches their experience with the cameras they are using. 

(2)
UM
Undisclosed Manufacturer #4
Jul 30, 2018

First of all, let's call them "estimators", not "calculators".  Also, from the comments, above, there is a risk of a conservative result negatively impacting sales.  Suggest the storage "estimator" produce a range of results..."low-middle-high", explaining the differences between best and worst case scenarios.

We once lost a project because our storage estimate was too high compared to a competitor.  Doing our lost-project analysis, we discovered that the spec called for 24 TB of storage. The spec also called for the storage to utilize RAID 5.  The competitor bid the entire drive system as available storage, whereas we increased the storage to provide the required net storage available under the RAID configuration. This caused us to include verbiage about RAID overhead in all future quotes.

Suggest any storage estimator include considerations for the various RAID overhead requirements. 

(2)
(3)
Avatar
Sean Patton
Jul 31, 2018

Great feedback and suggestion, thanks UM4

UM
Undisclosed Manufacturer #2
Jul 31, 2018

Agreed, but RAID overhead is largely predictable and deterministic. Basically add the drive for pairity.

Motion levels, quantization, true real world  smart codec efficiencies, et al are the "gotcha" that skew the results from one estimator to another.

It is very difficult to normalize storage calculations as much as it would be great to have a standard for all of us to adhere to.

That's the reason I believe a comparative estimator may be of value to IPVM subscribers.

(1)
New discussion

Ask questions and get answers to your physical security questions from IPVM team members and fellow subscribers.

Newest discussions